Data-Driven Investments at JGSB




  • Background courses

    • 665 Python for Business Analytics*
    • 649 Data Mining for Business Analytics
    • 679 Machine Learning for Business Analytics
  • Investments courses

    • 638 Data-Driven Investments Equity*
    • 639 Data-Driven Investments Credit*
    • 767 Data-Driven Investments Lab

Equity course

Inspiration

Data and procedures

  • 100+ monthly features used by GKX
    • financial ratios and growth rates
    • momentum, volatility, beta, market cap, volume
    • analyst forecasts and earnings surprises
  • Industry membership
  • Machine learning (random forests, boosted forests, neural networks) \(\rightarrow\) stock return predictions

= Optimized factor investing

Backtests

  • Long best stocks and maybe short worst stocks
  • Analyze portfolio returns
    • Sharpe ratios and drawdowns
    • CAPM alphas and information ratios
    • Fama-French factor attribution analysis
  • Group projects and presentations

Credit course

  • ML models for default prediction

Lab course

  • Full semester (spring)
  • Prereq is either equity or credit
  • More ML (cross validation, …)
  • New data sources (insider trades, …)
  • Implement models at Alpaca Brokerage using python API
  • Weekly team reports on performance and model evaluations

CQA

  • Several students taking the equity, credit, and lab courses are participating in the Chicago Quantitative Alliance investment competition.
  • Run a long/short portfolio with no position > 5% and cash < 5% and no ETFs.
  • Last year, a JGSB MBA team won the competition, finishing first in all three categories: return, compliance, and presentation.
  • This year, at the 2/3 point